Project-2 baseline

worker-toolkit-potion-polyglot/scripts/lib/patch-size.ts

@@ -0,0 +1,68 @@
+import { existsSync, readFileSync, statSync } from 'fs';
+
+const MB = 1024 * 1024;
+/** At or above this the package is refused: GitHub warns at 50 MB and rejects at 100 MB. */
+export const PATCH_MAX_BYTES = 49 * MB;
+export const PATCH_WARN_BYTES = 10 * MB;
+export const PATCH_BINARY_FILE_WARN_BYTES = 1 * MB;
+
+export interface PatchSizeReport {
+  bytes: number;
+  errors: string[];
+  warnings: string[];
+}
+
+function fmt(bytes: number): string {
+  return `${(bytes / MB).toFixed(1)} MB`;
+}
+
+/** Binary files in a `git diff --binary` patch whose size exceeds the threshold. */
+export function largeBinaryFiles(
+  patchText: string,
+  thresholdBytes = PATCH_BINARY_FILE_WARN_BYTES
+): { path: string; bytes: number }[] {
+  const out: { path: string; bytes: number }[] = [];
+  const lines = patchText.split('\n');
+  let current = '';
+  for (let i = 0; i < lines.length; i++) {
+    const line = lines[i];
+    const header = /^diff --git a\/.* b\/(.*)$/.exec(line);
+    if (header) {
+      current = header[1];
+      continue;
+    }
+    if (line === 'GIT binary patch') {
+      const m = /^(?:literal|delta) (\d+)$/.exec(lines[i + 1] ?? '');
+      const bytes = m ? parseInt(m[1], 10) : 0;
+      if (bytes > thresholdBytes) out.push({ path: current, bytes });
+    }
+  }
+  return out;
+}
+
+export function checkWorkspacePatchSize(patchPath: string): PatchSizeReport {
+  const report: PatchSizeReport = { bytes: 0, errors: [], warnings: [] };
+  if (!existsSync(patchPath)) return report;
+  report.bytes = statSync(patchPath).size;
+  if (report.bytes >= PATCH_MAX_BYTES) {
+    report.errors.push(
+      `environment/workspace.patch is ${fmt(report.bytes)} — the limit is ${fmt(PATCH_MAX_BYTES)}. ` +
+        'Remove large files (model weights, datasets, archives, build output) and regenerate the patch; ' +
+        'use a tiny dummy fixture or a stub instead.'
+    );
+    return report;
+  }
+  if (report.bytes >= PATCH_WARN_BYTES) {
+    report.warnings.push(
+      `environment/workspace.patch is ${fmt(report.bytes)} (limit ${fmt(PATCH_MAX_BYTES)}) — ` +
+        'check it contains only what the task needs'
+    );
+  }
+  for (const f of largeBinaryFiles(readFileSync(patchPath, 'latin1'))) {
+    report.warnings.push(
+      `workspace.patch adds a ${fmt(f.bytes)} binary file: ${f.path} — ` +
+        'prefer a tiny dummy fixture (e.g. a small random-init checkpoint) or a stub'
+    );
+  }
+  return report;
+}